I found it! in topaz, the output level was 0. (topaz 1> level 0). This
printed nil because all the instVars of our destroyed object is nil,
and it failed to print anything useful.
so,
level 1
display oops
run
Object _objectForOop: 967292673
%
showed that I had in deed an instance of CollectiveInvestmentScheme.
Apologies for the false alarm.
---------- Forwarded message ----------
From: Otto Behrens <
[hidden email]>
Date: Fri, Sep 25, 2015 at 5:48 PM
Subject: nil objects are found in listInstances:?
To: "
[hidden email]" <
[hidden email]>
Hi,
I just encountered a situation I find hard to understand. I'm running
a GS 3.2.6 and I'm doing some migrations. So after all my fixing and
things, I check if I find any "destroyed" objects. These are objects
in our domain, where we set a special instVar called #objectState to
#destroyed.
I do this query to check:
System abortTransaction.
(SystemRepository listInstances: {CollectiveInvestmentScheme})
do: [ :instances |
(instances select: [ :i | i isDestroyed ])
ifNotEmptyDo: [ :destroyedObjects | self error: 'More destroyed CIS
objects: ' , (destroyedObjects collect: [ :o | o asOop ]) printString
] ]
And, alas, I get the error: More destroyed CIS objects: anArray(
967292673, 968180993, 968574465, 968610561...
The weird bit is:
Object _objectForOop: 967292673
returns nil!
I run my query again, and find the same results.
SystemRepository objectAudit reveals no problems.
Please help! Am I smoking my socks?
Cheers
Otto
_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Locked
